Desktop search tools

[quant]

Originally posted by Quantum7
I have the "store in UR" option turned on by default. Therefore I really don't care which URL the original file is from (if it's not a web address) as I immediately afterwards delete the original file.

word of caution. I personally decided against storing inside UR (only the most relevant info is stored). While everything is keyworded, that's not enough and nothing special by today's standards. As your database will grow, with the current UR search capabilities, your search efficiency will get lower and lower.

What's missing is (IMHO) relevance search http://www.kinook.com/Forum/showthre...&threadid=1133

But the standard search engines offer much much more, I use SearchInform:
http://www.kinook.com/Forum/showthre...&threadid=2237
"it has fuzzy search, proximities, search through current results, look for synonyms, similar search, morphology and stem searches, number of found words, highlighting, etc... "

But if UR's database (with all the context, ie. all stored pdf,html,doc, ...) files is searchable through say SearchInform or other search engines ... that would be another story. I think this will become more and more urgent issue in the future as peoples databases grow (my db is already big enough to see the search deficiency of UR)

[Quantum7]

Thanks. Right now I only store directly PIM related documents like lists (roughly 15 excel sheets). And temporary documents (lists I have to convert to tasks / reference lists).

E-mail and other documents are in my archive's. These are so large (multiple GB's) that I have no intention of storing these in UR.

One application you might want to look at:
Archivarius, very cheap desktop search alternative with extensive capabilities and continuous development. I sent them an URD file as an example to support; no response yet though .

Link:
http://www.likasoft.com/document-search/

Note: Haven't looked (or heard of) Search Inform yet, sounds interesting.

[quant]

Originally posted by Quantum7
Note: Haven't looked (or heard of) Search Inform yet, sounds interesting.

SearchInform has my full recommendation. This is the free version, the limitation is that at any time only one index can be loaded http://www.searchinform.com/cgi-bin/...rmdeskfree.exe

For comparison
http://www.searchinform.com/search-s...sktop-buy.html

[quant]

Just an update, I gave two suggestions to Archivarius, to create a summary view, and advanced phrase search with word ordering ... seems that they will implement it in the next release. I'm going for it ... at a student price under $20, seems very good

Thanks for suggestion, Quantum7

[Quantum7]

If any mod reads this: Might want to split all posts dealing with desktop search engines to a general discussion thread .

Quant:
I'll probably compare both soon. So far Archivarius seems more powerful although I still have some features I'm missing there. Will have to test if those are present now (didn't try a new version for >6 months ).

[J-Mac]

Originally posted by quant
Just an update, I gave two suggestions to Archivarius, to create a summary view, and advanced phrase search with word ordering ... seems that they will implement it in the next release. I'm going for it ... at a student price under $20, seems very good

Thanks for suggestion, Quantum7

quant, are you saying that Archivarius can currently search your URDs? Or you are hoping that they implement this? Sorry, it is unclear to me.

And if not, can any current desktop search engine search UR's data?

Thank you,

Jim

[quant]

Originally posted by J-Mac
quant, are you saying that Archivarius can currently search your URDs? Or you are hoping that they implement this? Sorry, it is unclear to me.

no, it cannot. I think several people asked them if they could do it, but they didn't, probably too small user base for them to bother or maybe problem with urd file specification.
In my case, the documents that contain a lot of unprocessed info are only linked to UR (neither keyworded in UR). I keep in UR only the most important info. For the rest, Archivarius finds everything I look for, it's fantastic, plus very responsive support.

Originally posted by J-Mac
And if not, can any current desktop search engine search UR's data?

None that I know of. I think Kinook once said that they will "reinvestigate" to allow GDS indexing, but don't know the result. Kinook?

[J-Mac]

Thanks quant. Well, I won't even consider Archivarius at all unless I discover from someone else that it can do something great like search our URDs. Otherwise, I cannot trial the software to evaluate it because the developer limits his trial version to 10,000 files, which is not adequate for me (and several others I know) to test it. I can handle time-limited trials; that works fine for me. But limiting a search engine in how much it can search through? Unacceptable.

If you had said that it does indeed index the Ultra Recall databases I might have risked a purchase!

And as for GDS, I am afraid that won't help either. Google has a max limit that I apparently exceeded a long time ago. Currently GDS stops indexing on my machine at 72%. And even then it won't index a lot of my common file types.

I'll stick with alternating between X1 and Copernic with Locate32 as a filler.

Thanks!

Jim

[quant]

Originally posted by J-Mac
Thanks quant. Well, I won't even consider Archivarius at all unless I discover from someone else that it can do something great like search our URDs. Otherwise, I cannot trial the software to evaluate it because the developer limits his trial version to 10,000 files, which is not adequate for me (and several others I know) to test it. I can handle time-limited trials; that works fine for me. But limiting a search engine in how much it can search through? Unacceptable.

You could test it on part of your data, and it's a limitation per index (you can have many indexes, so in fact you could easily get around that limitation if you wanted), but fair enough. You're missing a lot, morphology, wild card support * ?, search for close words match, ...

[J-Mac]

@quant:

Sorry to get back to you so late. You may be right; it's just one of those annoyances that nagged me back last year when I did trial it. There was no notification of that limitation on the web site at that time, and I didn't realize then why it was only indexing such a small part of my data until I asked about it!

And another thing that ticked me off with GDS: not too long after it was released and I was using it, I received a notice from Google that my account, along w/ several, had inadvertently been made available on the web and also made public for a short time. Yikes! No indication that it was seen by anyone, but my account was supposedly configured so that even I could not see my own desktop index during a web search - I felt that was possibly putting my data at risk. Ha! So they accidentally made it public anyway!

That's when I decided to make it a personal rule to never allow my personal data to be entrusted with a "free" application from a company best known for the largest internet search engine again. (My Gmail account is used strictly for forum notifications and certain newsletters. I use a paid email account for all else.)

Maybe I'll stop fussing about the trial stuff and give Archivarius a shot after all.

Thanks quant.

Jim

[MrAnalogy]

Originally posted by quant
Just an update, I gave two suggestions to Archivarius, to create a summary view, and advanced phrase search with word ordering ... seems that they will implement it in the next release.

UR is not listed among the formats it can read:
http://www.likasoft.com/document-search/features.shtml

Do you know if it supports UR?

[quant]

Originally posted by MrAnalogy
UR is not listed among the formats it can read:
http://www.likasoft.com/document-search/features.shtml

Do you know if it supports UR?

it doesn't. On the UR roadmap, there is "Investigate desktop search integration again" ... so maybe something has to be done from Kinook's side first?

[TMF]

Originally posted by quant
it doesn't. On the UR roadmap, there is "Investigate desktop search integration again" ... so maybe something has to be done from Kinook's side first?

I believe the investigation is more about other desktop search engines like the Google one, where Kinook would have to create a custom plugin for GDS.
For Archivarius IMHO no steps from Kinook side are needed, file is sqlite database and the only problems are that it has slightly modified header (which shouldn't be a biggie), and that passworded databases and binary data in the databases wouldn't be accessible to Archivarius.

[quant]

Originally posted by TMF
For Archivarius IMHO no steps from Kinook side are needed, file is sqlite database and the only problems are that it has slightly modified header (which shouldn't be a biggie)

heh, interesting ... so isn't sqlite a biggie for GDS if it needs a plugin? These big monopolies have it nicely thought out, instead of them doing their job, you need to do the job if you want to be "in".